PixelDAQ PM, 16-4-2004 ---------------------- Remarks on PixLib by Paolo beforehand (meeting shared by Tom Meyer): ==================================================================== PixLib ------ I've put in the repository a new version which can handle occupancy histograms from a threshold scan. You also find a new example (ScanTest) showing the histograms via a simple root interface. I will send an instruction message later. Tobias already committed the first version of PixBoc. To use it however we have to add few new methods in PixController. Tobias sent me the updates and I will integrate them as soon as I can. Guido is adding Histo support to RootDB and when this is ready ScanTest will be able to save the histograms in a root file. Carlo is updating PixMCC and PixFE to the same configuration scheme already used for PixScanConfig. The stability of the different components is: PixConfDBInterface: interface stable; histogram handling being implemented. PixController: interface almost stable after the modif for boc. Multi module functionalities still untested PixScanConfig: the idea I'm following here is to add to this class the results of the scan, maybe having two different methods for reading/writing from/to db, one including histo and the other limited to the scan configuration. This object can also be used as a pure memory container of scan config and result. If we keep this direction we have to extend the interface to include any type of histo/fit being produced by the rod. Otherwise we have to crate a new container (PixScan?) and use it to store a PixScanConfig and the histograms. If you take a decision I can implement it. PixRunConfig: this is just an interface. The design should be similar to the one of PixScanConfig (here we will have to store the rod monitoring histo). PixModuleGroup: very often this class is bypassed in the examples and PixController is used directly. As a consequence the interface is out of date. I will try to fix it next week. Local implementation for non intelligent controllers still missing. Not a priority. In part is being implemented by Markus for NA60. PixModule: Same as PixModuleGroup. PixFe: config parameters being re-arranged. PixMcc: config parameters being re-arranged. PixDcs: initial version; interface agreed in Wuppertal. PixBoc: initial functional version; interface complete? Release ------- More than for the delay itself, I'm worried by the fact Tom is leaving next week, and I don't know if he managed to test the new version of RodModule with the new DSP code. If not we may have problems when the new DSP is released. I propose to try a release of PixLib+STcontrol by the end of next week with scanning+histogramming+gui+(at least two modules) based on the current release of the DSP code. This will be not very useful for system test being limited to FEI1, but will be ok for the coordination between PixLib and STcontrol. As soon as the new DSP code is available we start testing it with the new RodModule and we try to make PixLib working again. If we manage to do so by the end of next week we're lucky, otherwise we release first for the old DSP code and then update for the new version. ------------------------------------------------------------------------------- Date: Fri, 16 Apr 2004 11:04:25 -0500 From: W T Meyer To: Atlas-Pixel-Daq Subject: Summary of today's meeting Things are starting to shape up. Here are the important conclusions we reached: 1. Douglas will get DSP code out today and it will be put in the repository, along with the latest version of CommonWithDsp. 2. Tom will commit the new RodCrate and RodUtils files at the same time. 3. These will not be tagged yet. 4. PixLib and STcontrol will be ready by the end of next week. 5. On Friday, April 23, we will tag the repository as Pixel-01-00. (Paolo, I will not be here, can you do the tagging, or see that someone does it?) We will have a phone meeting as usual next Wednesday at half past whatever hour it is where you are. Cheers,